Threading instability, simultaneous saturate (no canceled tasks)

Nobody has claimed this yet.

So in a homogeneous ring, there are times when it would be nice to check if an ideal contains a power of the irrelevant. I was thinking maybe try to saturate with respect to each variable individually. Indeed, in an example I was playing with, saturating with respect to each variable separately (8 variables) takes about 0.5 seconds per variable, but saturating with respect to the maximal ideal takes about 6 seconds.

So I decided to write something to see if could take advantage of multithreading here.

```
threadedSaturate = method();

threadedSaturate(Ideal, List) := (I1, L1) -> (
taskList := apply(L1, z -> createTask(saturate, (I1, ideal(z))) );
apply(taskList, t -> schedule t);
while true do (
nanosleep 1000000; --one thousandth of a second
if (all(taskList, t -> isReady(t))) then break;
);
resultList = apply(taskList, t -> taskResult(t));
return resultList
);
```
Unfortunately, it frequently (but not always) gives errors (which become worse the higher `allowableThreads` is).

For example, `allowableThreads = 4` sometimes gives me things like (notice it looks like one of the threads crashed and didn't return an ideal).
```
elapsedTime threadedSaturate(J3, first entries gens mm)
/usr/share/Macaulay2/Core/intersect.m2:41:24:(1):[23]: error: intersect: expected at least one object
/usr/share/Macaulay2/Core/methods.m2:105:83:(1):[22]: --back trace--
/usr/share/Macaulay2/Core/option.m2:16:8:(1):[21]: --back trace--
/usr/share/Macaulay2/Core/methods.m2:105:83:(1):[19]: --back trace--
/usr/share/Macaulay2/Saturation.m2:495:14:(2):[17]: --back trace--
/usr/share/Macaulay2/Core/methods.m2:559:51:(1):[16]: --back trace--
/usr/share/Macaulay2/Core/methods.m2:573:23:(1):[15]: --back trace--
/usr/share/Macaulay2/Core/methods.m2:572:26:(1):[14]: --back trace--
/usr/share/Macaulay2/Core/option.m2:38:8:(1):[13]: --back trace--
/usr/share/Macaulay2/Core/methods.m2:119:80:(1):[12]: --back trace--
/usr/share/Macaulay2/Core/option.m2:38:8:(1):[11]: --back trace--
/usr/share/Macaulay2/Core/methods.m2:565:83:(1):[10]: --back trace--
/usr/share/Macaulay2/Core/option.m2:38:8:(1):[9]: --back trace--
/usr/share/Macaulay2/Core/methods.m2:119:80:(1):[8]: --back trace--
/usr/share/Macaulay2/Core/option.m2:38:8:(1):[7]: --back trace--
/usr/share/Macaulay2/Saturation.m2:418:9:(2):[6]: --back trace--
/usr/share/Macaulay2/Saturation.m2:140:23:(2):[5]: --back trace--
/usr/share/Macaulay2/Saturation.m2:425:10:(2):[4]: --back trace--
/usr/share/Macaulay2/Saturation.m2:367:62:(2):[3]: --back trace--
/usr/share/Macaulay2/Core/methods.m2:119:80:(1):[2]: --back trace--
/usr/share/Macaulay2/Core/option.m2:16:8:(1):[1]: --back trace--
/usr/share/Macaulay2/Core/option.m2:16:8:(1): --back trace--
-- 1.50479 seconds elapsed

o44 = {, ideal 1, ideal 1, ideal 1, ideal 1, ideal 1, ideal 1, ideal 1}

o44 : List
```
On the other hand, it is nearly a linear increase in speed (4 threads is 1.5 seconds).

Start with the saturation paths in Saturation.m2 and the intersect call reported in Core/intersect.m2, then trace how createTask, schedule, isReady, and taskResult handle failures or missing results. Reproduce the threadedSaturate example with different allowableThreads values and verify that every task returns an ideal without errors.

Written by the indexing model from the issue text.